In the northern cereal belt of Australia, farmers are reluctant to apply nitrogen (N) fertilizers because of a perception that if N is added to the soil and no crop is subsequently planted due to lack of rain, the N is 'lost'. An experiment was conducted on a cracking clay soil over three seasons to compare the response of grain sorghum to N applied to the current crop v. N applied the previous season which was then either planted or left fallow (to simulate a missed planting opportunity). Recovery of 15~-labelled fertilizer by the crop and that remaining in the soil were simultaneously determined in microplots. The effect of tillage practice [zero (ZT) and conventional (CT)] was also examined. Sorghum grain yield responded to fresh applications of N in 1993 and 1993194 but not 1992, reflecting the importance of timing of rainfall rather than the total amount received within the season. Applications of N to the current crop always improved yield more than equivalent amounts of N applied to the previous crop. Grain yields of plots that were previously fallowed (fallow-sorghum rotation) were higher than the combined yields of sorghum-sorghum rotations, although fallowing was an inefficient means of accumulating both water and mineral N. Recovery of applied 1 5 ~ by sorghum varied from 48% in 1992 to 36% in 1993 but was not related to the overall N responsiveness of the crop. Sorghum recovered a similar proportion of 1 5 ~ from plots which had been fertilized and then fallowed the previous year compared to fresh applications to the current crop, despite the fallow plots having less 1 5 ~ in them due to losses from the previous season. Losses of 1 5 ~ from the soil/plant system varied markedly with year and appeared to be related to the pattern of rainfall occurring and its possible effect on denitrification. Tillage practice did not affect grain yields or PAWC, had minimal effect on the amount of mineral N present, and little influence on the fertilizer N requirements of sorghum per se. This study suggests that there is only a small residual value to subsequent sorghum crops of fertilizer N if added initially to a successful crop. However, if N is applied pre-plant and the crop is not planted, for example due to lack of planting rain, a large proportion of this N can remain available to the following crop depending on the nature of the subsequent rainfall.